The mechanism by which probiotic lactobacilli affect the immune system is strain specific. As the immune system is a multicompartmental system, each strain has its way to interact with it and induce a visible and quantifiable effect. This review summarizes the interplay existing between the host immune system and probiotic lactobacilli, that is, with emphasis on lactobacilli as a prototype probiotic genus. Several aspects including the bacterial-host cross-talk with the mucosal and systemic immune system are presented, as well as short sections on the competing effect towards pathogenic bacteria and their uses as delivery vehicle for antigens.
The concept of probiotics probably dates back to 1908, when Noble Prize Winner Eli Metchnikoff suggested that the long life of Bulgarian peasants resulted from their consumption of fermented milk products. The term ‘probiotic’ was first used in 1965, by Lilly and Stillwell for describing substances secreted by one organism which stimulate the growth of another (Gupta and Garg 2009). The term was derived from the Greek word, meaning ‘for life’ (Reid et al. 2003). In a broadened definition, Naidu et al. (1999) describes probiotics as microbial dietary adjuncts that beneficially affect the host physiology by modulating the immune system as well as improving nutritional and microbial balance in the intestinal tract. An expert panel commissioned by FAO (Food and Agriculture Organization of the United Nations) and WHO (World Health Organizations) defined probiotics as ‘live micro-organisms’ which, when administered in adequate amounts, confers a health benefit on the host (FAO/WHO 2001). Marteau et al. (2002) defined them as ‘microbial preparation or components of microbial cells that have a beneficial effect on health and well-being’. Nowadays, the mainly accepted definition states probiotics should be live micro-organisms.
The immune system can be divided into two general arms: innate (natural or nonspecific) and adaptive (acquired or specific) immunity, which work together synergistically (Cummings et al. 2004). In mammals, it has an extremely sophisticated adaptive immune arm of both systemic and mucosal (local) type, where a complex array of cells and molecules interacts to provide protection from challenge by pathogenic microbes (bacteria, viruses, parasites). In vitro experiments and studies with germ-free animals have convincingly demonstrated that a healthy intestinal microbiota mediates important roles in normal gut regulation and homeostasis, influencing epithelial growth and survival, innate and adaptive immune development and regulation, competitive exclusion of pathogens (Bienenstock et al. 2013). During their interaction with the host, probiotics forming microbiota perform in different manners and affect several parts of the host immune system. Therefore, the use of Lactobacillus as probiotic for human requires fine-tuned selection of biologically relevant strains and a comprehensive understanding of its action on immune system elements. This review explores in a multidimensional way, how probiotic especially lactobacilli strains stimulate mucosal and systemic immune system to thoroughly elucidate the mechanism of action.
Mucosal immune system
The mucosal surfaces include the surface lining of the gastrointestinal, respiratory and genitourinary tracts, where the encounters with antigens or infectious agents occur (Delves and Roitt 2000). Mammals have developed an extremely sophisticated adaptive immune system of both systemic and mucosal (local) type. The immune properties of the digestive mucosa are provided by the GALT (gut-associated lymphoid tissue), composed of lymphoid aggregates, including the Peyer's patches (located mainly in the small intestinal distal ileum), where induction of immune response occurs, and mesenteric lymphoid nodes (Izcue et al. 2009). The mucosal immune system also consists of physical (mucus), molecular (antimicrobial proteins) and cellular components that act synergistically to prevent microbes from invading the body. Epithelial cells have long been understood as a primary physical barrier, forming a selective cellular barrier, and are now recognized as a principal interface with the microbiota, initiating immune responses (Koch and Nusrat 2012). The probiotics are clearly involved in the anatomic and functional development of mucosal immunity mediated by intraepithelial lymphocytes and other immunomodulatory cells (cytokine-producing cells, phagocytic cells, goblet cells, IgA-secreting cells) resident in the mucosal lamina propria (Fig. 1). Many probiotic strains, upon adherence and colonization of the gut, (Lebeer et al. 2008) can manipulate the host's mucosal immune system through secretion of ‘immunomodulins’ that affect cell signalling pathways in mammalian cells (Hemaiswarya et al. 2013). Some of these effects are direct on immune cells such as dendritic cells (DC) which can themselves protrude through intact epithelial tight junctions and sample luminal contents and even engulf bacteria. Using an intestinal loop mouse model and fluorescently labelled bacteria, it was proven that entry of commensal or probiotic bacteria in Peyer's patches (PP) via the microfold cell pathway was mediated by their association with S-IgA (Rol et al. 2012). Other probiotic bacteria promote the generation and up-regulation of regulatory DC and T cells and this is associated with increased production of the regulatory cytokines TGFβ and especially IL-10 (Bienenstock et al. 2013). This part will delineate the effects of lactobacilli on different compartments of intestinal mucosal immune system.
Effects on secretory-IgA levels and IgA-secreting cells
In the lamina propria of the gut, B cells differentiate into plasma cells and secrete dimeric IgA antibodies. The polymeric Ig receptor (pIgR) on the basolateral surface of intestinal epithelial cells complexes with IgA and transports it to the apical cell surface where it is secreted into the intestinal lumen (Peterson et al. 2007). Secretory IgA is multifunctional in mucosal-associated immunity. It helps protecting the host by binding a variety of antigens (bacteria, virus, fungi, etc.) thus avoiding opportunistic pathogens to enter and disseminate in the systemic compartment (a process known as immune exclusion). In addition, it assists in controlling the necessary symbiotic relationship existing between commensals and the host (Corthesy 2013). S-IgA fulfils its function at mucosal surface by sampling of antigen-S-IgA complexes by microfold (M) cells, intimate contact occurring with Peyer's patch dendritic cells (DC), down-regulation of inflammatory processes, modulation of epithelial cells and clearing of immune complex by peristalsis (Corthesy 2013).
There is evidence that probiotic lactobacilli affect the IgA secretion in intestinal fluid and the number of IgA-secreting cells. For instance, Lactobacillus casei Zhang (Ya et al. 2008) and Lactobacillus crispatus KT-11 (Tobita et al. 2010) have shown the ability to improve the gastrointestinal mucosal resistance to infections by inducing a substantial increase in the concentration of S-IgA in intestinal fluid in a dose-dependent manner compared with control. Similarly, the number of IgA-secreting cells in lamina propria of small intestine increased significantly 7 days after the administration of Lact. casei CRL 431, but the intestinal fluid did not contain specific S-IgA against the probiotic culture epitopes, suggesting that there was no antigen presentation of Lact. casei epitope to Th-2 cells after 7 days of administration (Maldonado and Perdigon 2006). Even after a long-term (98 days) oral administration of fermented milk containing Lact. casei DN-114001, Lactobacillus delbrueckii ssp. bulgaricus and Streptococcus thermophilus, the numbers of IgA+ cells in both the small and large intestine increased (De Moreno de LeBlanc et al. 2008) as well as in mucosal sites distant to the gut such as bronchus and mammary gland, by migration of the IgA+ cells to the mesenteric lymphoid node and then via the thoracic duct to the circulation, arriving in the bronchus and mammary glands (Galdeano et al. 2007; De Moreno de LeBlanc et al. 2008). The mechanism underlying is not yet clear but it has been proven that epithelial cells and DCs, under stimulation of probiotic or microbiota, produce molecules such as A Proliferation-Inducing Ligand (APRIL), CD40 ligand and TGF-β that induce T cell-independent IgA class switching (Peterson et al. 2007). Lactobacillus plantarum NCIMB88 and Lact. plantarum WCFS1 induced Intestinal Epithelial Cells (IECs) to produce APRIL, which triggers IgA class switching to IgA2, an immunoglobulin more resistant to bacterial protease (Peterson et al. 2007).
Effects on cytokine-producing cells
The most fascinating feature and extensively studied effect of immune system modulation by probiotic bacteria is cytokine production. Probiotic bacteria induce the secretion of cytokines from IEC in a strain-specific manner. Moreover, the same cytokine can be produced by multiple cell types; can have multiple effects on the same cell and act on many different cell types (Ashraf and Shah 2014). IECs are the initial point of contact between the host and intestinal microbes and they communicate extensively with commensal bacteria and probiotics, influencing the inflammatory signalling pathway in IECs. The key signalling channel is the nuclear factor-kappa B (NF-ĸB) pathway. Under nonstimulatory conditions, NF-κB is present in its inactive form in the cytoplasm, bound to the inhibitor molecule, the inhibitor of NF-κB (IκB). When pro-inflammatory stimuli trigger signalling pathways, IκB is phosphorylated by IκB kinase (IKK), targeting it for ubiquitination and subsequent proteasomal degradation. Once freed from IκB, NF-κB is able to migrate into the nucleus, bind target promoters and activate transcription of effector genes like pro-inflammatory genes (Thomas and Versalovic 2010). Several probiotic strains can prevent degradation of IκB, therefore preventing the expression of pro-inflammatory cytokine like IL-8 by IEC (Thomas and Versalovic 2010). A study by Zhang et al. (2005), investigating the effects of both viable and heat-killed Lactobacillus GG in an epithelial cell model, demonstrated the probiotic's ability to decrease IκB degradation and subsequent NF-κB translocation into the nucleus, resulting in decreased TNF-induced IL-8 production (Zhang et al. 2005). Others have shown that pretreatment of epithelial cells with Lact. casei DN-114 001 decreased Shigella flexneri-induced NF-κB activation due to inhibition of IκBα degradation (Tien et al. 2006).
The anti-inflammatory role of IL-6 in the enhancement of IgA secretion has been well demonstrated. This cytokine produced by IEC, macrophages and T cells, has the ability to induce the terminal development of B cells in plasmatic cells, which express IgA (Goodrich and McGee 1999). Maldonado and Perdigon (2006) found an intercorrelation between the increase in IL-6-producing cells and IgA-secreting cells in lamina propria of mice 7 days after administration of Lact. casei CRL 431. Mast cells also stimulated by probiotic, can produce the cytokine IL-4, which together with IL-6 and TGF-β, induce the T-independent switch from IgM to IgA on the surface of B lymphocytes, thereby enhancing the production of IgA (Galdeano et al. 2007). The preferential switching to IgA could also be explained by the fact that bacterial and food-derived products present in the gut (lipopolysaccharide and retinoic acid) condition Follicular DCs and facilitate the activation of TGF-β1 and secretion of large amounts of B-cell activating factor (BAFF) (Suzuki et al. 2010; Macpherson et al. 2012), the two essential cytokines that direct switching towards IgA and survival of recently switched IgA+ B cells, respectively (Kato et al. 2014). In a recent study, a low dose (106 CFU ml−1) of Lactobacillus acidophilus NCFM significantly promoted IFN-γ-producing T cell responses and down-regulated T-regulatory (Treg) cell responses and their TGF-β and IL-10 productions in intestine and spleen, compared with the high dose (109 CFU ml−1) and control groups (Wen et al. 2012a). Conversely, high dose increased the frequency of Treg cells in most of the tissues compared with the control groups (Wen et al. 2012a). This study therefore suggests that Lact. acidophilus NCFM can be ineffective or even detrimental if not used at the optimal dosage for the appropriate purposes. Further work needs to be performed on other probiotics to determine more fully the effects of different doses.
Dendritic cells (DCs) directly sample gut luminal contents through cellular processes that extend between IEC into the lumen. This feature of intestinal DCs, combined with their ability to orchestrate T-lymphocyte responses, highlights the role of DCs as a bridge between microbes, innate immunity and adaptive immunity (Thomas and Versalovic 2010). Intestinal DCs can harbour viable commensal bacteria and transport the bacteria to the mesenteric lymph nodes (MLNs) through M Cells, where it is retained and prevented from entering the systemic immune compartment (Rochereau et al. 2013). By this mechanism, DCs can selectively induce IgA to protect against mucosal invasion, while restricting immune responses to the local microbial community in the gut (Macpherson et al. 2005).
It has been shown that different species of Lactic Acid Bacteria (LAB) possess the ability to finely regulate DC maturation and their interactions with Natural Killer (NK) cells, polarizing the subsequent T cell activity towards Th1, Th2 or even Treg responses (Fink et al. 2007). This network can, therefore, control not only the strength and the quality of innate responses, but also the subsequent adaptive responses, via both cell-to-cell interactions and cytokine release (Rizzello et al. 2011). A study on the dose-dependent immunomodulation of human DCs by the probiotic Lactobacillus rhamnosus Lcr35 showed that very different profiles of gene expression were induced with different doses, displaying pro- or anti-inflammatory profiles (Evrard et al. 2011). The glycosylation of the S-layer protein produced by Lact. acidophilus NCFM is considered to be a key step for its interaction with the C-type Lectin Receptors DC-SIGN (DC-specific ICAM3-grabbing nonintegrin) and influences cytokine response in DCs and T cell priming (Konstantinov et al. 2008).
The impact of bacterial growth stage on host communication has been clearly demonstrated by the substantially different effects elicited in the intestine mucosa by exponentially growing vs stationary phase derived from Lact. plantarum bacteria. The Lact. plantarum preparations derived from the stationary phase of growth have been found to stimulate prominently immune pathways, although this has not been observed by logarithmically growing bacteria of the same strain (van Baarlen et al. 2009). Some probiotic strains may also stimulate regulatory immune mechanisms by the generation of regulatory DCs and CD4+Foxp3+ T cells (Kwon et al. 2010). However, it remains unclear whether the mechanism involves Microbe-Associated Molecular Pattern (MAMP)-mediated Toll like Receptor (TLR) signalling on Treg cells, DCs or IECs. But after their activation by Antigen Presenting Cells (APCs), Treg cells migrate in MLN via lymphatics, presumably in an S1P receptor 1-dependent process and feedback on immune induction at the systemic level (Pabst 2013).
Effects on TLRs expression in lamina propria and Peyer's patches
Another fundamental property of the epithelia is its ability to monitor bacterial presence, which is generally recognized to involve host perception of prokaryotic macromolecular motifs (such as LPS, peptidoglycan and flagellin) that are bound and recognized by pattern recognition receptors (PRRs), such as TLRs and Nod (Nucleotide oligomerization domain) proteins (Pedron and Sansonetti 2008). These macromolecules, designated MAMPs, are generally invariant structural components of the bacteria, including specific cell wall polysaccharides, peptidoglycan (PG), lipoprotein anchors, lipoteichoic acids (LTA) anchored in the cytoplasmic membrane, wall bound teichoic acids (WTA) and polysaccharides (Van Baarlen et al. 2013). MAMP bound PRRs result in the activation of signalling pathways eventuating in host gene regulatory events that likely have broadly cytoprotective effects and with quantitative or qualitatively increased stimulation (Bienenstock et al. 2013) including immune activation, antigen presentation and expression of antimicrobial factors (Lebeer et al. 2010; Wells et al. 2011; Bron et al. 2012a).
The TLRs are a family of single-pass transmembrane glycoproteins, which contain 10 members in humans and 13 in mice (Lavelle et al. 2010). TLR1–TLR7 and TLR9 have been characterized to recognize microbial components. In signalling pathways via TLRs, a common adaptor, MyD88, was first characterized as an essential component for the activation of innate immunity (Dunzendorfer et al. 2004; Takeda and Akira 2004). MyD88 is essential for the induction of inflammatory cytokines triggered by all TLRs through NF-kB or MAP Kinase pathways (Takeda and Akira 2004). TLRs stimulation by probiotics has been widely studied in vivo and in vitro with different types of immune cells. The intensity of stimulation appears to be strain specific. Lact. casei CRL431 administrated to mice for 2, 5 or 7 days induced a significant increase in the number of TLR-2+ cells in the lamina propria of small intestine, with the enhancement being greater for 7 days postadministration. The effect in the number of TLR-2+ cells in Peyer's patches was enhanced only after 5 and 7 days postadministration (Maldonado and Perdigon 2006). Lactobacillus paracasei ssp. paracasei DC412 exhibited strong immunoregulatory activity and interacted strongly with gut-associated lymphoid tissue (GALT) of BALB⁄c inbred mice or Fisher-344 inbred rats through stimulation of TLR2⁄TLR4-mediated signalling events leading to secretion of a certain profile of cytokines namely, IFN-γ, TNF-α, IL-6 and IL-10 (Kourelis et al. 2010). TLR-2 recognizes a variety of microbial components such as lipoproteins/lipopeptides from various pathogens, PGs and LTA from Gram-positive bacteria (Dunzendorfer et al. 2004). TLR2 receptors mainly recognize the PG which is the principal component of the Gram (+) bacteria such as Lactobacillus genus and the ligands bound TLR2 increase the signals to produce cytokines, but also to strengthen the epithelial barrier (Abreu et al. 2005). TLR4 recognizes the LPS present in the cell wall of the Gram (−) bacteria. Upon activation, it initiates an innate immune response leading to the induction of pro-inflammatory mediators, the increase in TLR2 expression and the reduction in its own expression, which leads to the recruitment of inflammatory cells and the initiation of the appropriate responses in the spleen (Weiss et al. 2004). Flagellated bacteria like Salmonella can interact with TLR5 to induce activation of pro-inflammatory gene programmes for host protection (Fournier et al. 2009).
The immunostimulating effect of probiotic lactobacilli can lead to a different pattern of modulation of TLRs as they are lived or dead. For example, live Lact. casei Zhang (LcZ) promotes TLR2 mRNA transcription, whereas heat-killed LcZ enhances transcription of TLR2, TLR3, TLR4 and TLR9 in the murine macrophage cell line RAW264·7 (Wang et al. 2013). Using the human embryonic kidney NF-κB reporter cells expressing TLR-2/6 exposed to purified WTAs and/or the teichoic acid (TA) mutants from Lact. plantarum WCFS1, it was found that WTA is not directly involved in TLR-2/6 signalling, but attenuates this signalling in a backbone-independent manner, likely by affecting the release or exposure of immunomodulatory compounds such as LTA (Bron et al. 2012b), highlighting a shielding effect on PRRs reported elsewhere after activation by WTAs and polysaccharides (Remus et al. 2012).
The reason of strain specificity of probiotics could be laid in the chemical structure variation of MAMPs in terms of polymer composition, length and substitutions. Exemplary to this notion are the studies in several lactobacilli that have targeted the role of D-alanylation of LTA in the immunogenicity of these MAMPs. Loss of D-alanylation of LTA in Lact. plantarum by mutation on dlt operon decreases the capacity of the molecule to elicit pro-inflammatory responses in a TLR2-dependent manner (Grangette et al. 2005). In a mouse model of colitis, the dlt mutant was significantly more protective than its wild-type counterpart (Grangette et al. 2005). Moreover, the species-specific anti-inflammatory effects could be assigned to a particular, NOD-2 binding, PG-associated muropeptide that is released into its environment by protective Lactobacillus salivarius Ls33 but not by the nonprotective Lact. acidophilus NCFM strain (Macho-Fernandez et al. 2011). In addition, it has been shown that variations in the chemical modification (acetylation or pyruvylation) of the conserved PG polymer backbones may contribute to strain- or species-specific immunomodulatory capacities (Kleerebezem et al. 2010).
Effects on negative regulators of TLRs signalling pathways
Although strong pro-inflammatory responses after TLRs stimulation may be beneficial, TLR signalling needs to be fine-tuned to achieve an effective response against dangerous exogenous and endogenous factors, limiting a pathological reaction and restoring the immune balance. Therefore, several mechanisms have evolved which negatively regulate TLR-induced cellular responses at multiple levels of the signalling cascade (Gomariz et al. 2010). These mechanisms act at different levels: as extracellular regulators (soluble TLR2 and TLR4), as transmembrane protein regulators (RP105, ST2L, single immunoglobulin interleukin-1-related receptor (SIGGIR), TNF-related apoptosis-inducing ligand receptor (TRAILR)), as negative regulators of adapter molecules (soluble MyD88 (MyD88s), sterile α and heat-armadillo motifs (SARM), suppressor of cytokine signalling 1 (SOCS1)), as factors affecting effector proteins (interleukin-1 receptor-associated kinase M (IRAK-M), IRAK2, TNF receptor-associated factor 4 (TRAF4), toll-interacting protein (TOLLIP), A20, β-arrestins 1 and 2, LIND, FLN29, CYLD, SHIP-2) and finally as molecules affecting transcription factors (Pin-2, phosphatidylinositol 3-kinase (PI3K), ATF-3, Bcl3) (Liew et al. 2006). There are few reports highlighting the effects of probiotic lactobacilli on negative regulators of TLRs. The pretreatment of Lact. plantarum genomic DNA inhibited the phosphorylation of MAPKs and NF-kB, and also inhibited LPS-induced TNF-α production in response to subsequent LPS stimulation (Kim et al. 2012). Lactobacillus plantarum genomic DNA-mediated inhibition of signalling pathway and TNF-α was accompanied by the suppression of TLR2, TLR4 and TLR9 and the induction of IRAK M, a negative regulator of TLR (Kim et al. 2012). Lactobacillus jensenii TL2937, a strain with a high capacity to activate TLR2, was also the strain with the highest capacity to down-regulate IL-6 and IL-8 production by porcine intestinal epitheliocyte cell line (PIE cells) in challenge with enterotoxigenic Escherichia coli (ETEC) and LPS probably by inhibiting NF-κB and MAPK signalling pathways (Shimazu et al. 2012). Moreover, it was found that negative TLRs regulator MKP-1, A20 and Bcl-3 mRNA expression was up-regulated in PIE cells stimulated with Lact. jensenii TL2937 consequently limiting a pathological reaction (Shimazu et al. 2012). Some probiotics activate anti-inflammatory and regulatory immune effects in the settings of enteric infections and mucosal inflammation. Lactobacillus paracasei CNCM I-4034 and its supernatant dramatically reduce the production of IL-6, IL-8, IL-12p70 and TNF-α in human intestinal DCs challenged with Salmonella typhi (Bermudez-Brito et al. 2012). These authors demonstrated that Lact. paracasei CNCM I-4034 activates the expression of TLR2 in DCs, up-regulates the expression of TOLLIP and promotes the stimulation of TGF-β2, whereas the supernatant of the probiotic increases the secretion of TGF-β1, highlighting the importance of PRRs for probiotic actions (Bermudez-Brito et al. 2012). In another study involving Lact. jensenii TL2937 and swine DC CD172a+, the expression of A20 and TGF-β was not abolished by treating with anti-TLR2 antibodies, suggesting that other PRRs, in addition to TLR2, may be involved in the anti-inflammatory effect of Lact. jensenii TL2937 on CD172a+ APCs (Villena et al. 2012).
Effects on villi structure, mucin and antimicrobial peptide production
The structure of intestinal villi, the number of goblet cells involved in antibacterial production and the incidence of mucin-producing cells are of importance when looking for the health status of intestine. In a study performed to evaluate the impact of probiotic fermented milk with Lact. casei DN-114-001 (108 CFU ml−1) in nonsevere protein-energy-malnutrition mice model, all the test groups showed similar values for goblet cells without significant differences compared with both control groups (malnourished and well-nourished) (Galdeano et al. 2011). The intestinal villi of malnourished control group were short and their numbers were lower than in the well-nourished and Lact. casei group (Galdeano et al. 2011).
Goblet cells synthesize and release mucins, which are highly glycosylated proteins. Mucins are either secreted or membrane bound, and form a gel layer which acts as a protective barrier over the epithelial cells to protect from the harsh lumen environment (McClemens 2011). Lactobacillus crispatus K-11 increased mucin 13 and defensin alpha genes expression in Peyer's patches by twofold compared with control, ensuring constant protection against the attack of digestive fluid, micro-organisms, pollutants and toxins (Tobita et al. 2010). Moreover, the pili-encoding operon spaABC of Lact. rhamnosus GG (LGG) coding for the SpaC protein has been shown to bind to mucus, providing an explanation for the longer persistence of this strain in the human intestine compared with a closely related Lact. rhamnosus strain that lacks pili (Kankainen et al. 2009). The spa operon-encoded pili have also been proposed to modulate IL-8 induction in Caco-2 cells (Lebeer et al. 2012).
In the intestinal fluid, there are two main types of antimicrobial peptides: cathelicidins and defensins. The former is constitutively expressed by IECs while defensins are expressed primarily by Paneth cells (α-defensin) and IECs (β-defensin) (Wehkamp et al. 2004; Kelsall 2008). The effect of probiotic lactobacilli on cathelicidin production was not investigated up to date, may be because of its permanent expression. In the investigation of Akbari et al. (2008), it was found that a probiotic mixture had no effect on cathelicidin gene expression in chicken tonsils after 5 days of feeding (Akbari et al. 2008). Conversely, in vitro studies have shown that β-defensins expression or secretion was significantly up-regulated in Caco-2 cells upon stimulation by several lactobacilli species and VSL#3 (Schlee et al. 2008). Additionally, it was demonstrated that flagellin is a major stimulatory factor of defensin expression (Schlee et al. 2007) and defensin secretion required MAP Kinases, ERK ½, p38 and JNK pathways in vitro (Wehkamp et al. 2004; Schlee et al. 2008).
Systemic immune system
The systemic response of immune system encompasses cytokine and immunoglobulins produced in serum, blood lymphocyte, monocyte and macrophage responses and spleen lymphocyte stimulation. It is not confined to the initial infection site, but works throughout the body. Systemic responses can result from sampling of microbial antigens by DCs in intestinal lumen (usually under epithelial M cells). DCs can transport antigens to local lymph nodes, present antigens to naive immune cells and activate disparate effector responses from B, T helper (Th) and Treg cells, which will further initiate a release of a distinct panel of cytokines that are capable of delivering activating and inhibitory feedback signals to effector immune cells (Hord 2008). The following sections briefly describe interactions between probiotic lactobacilli and parts of systemic immune system above cited with suitable examples.
Effects on cytokines produced by T-lymphocyte subsets
Cytokines are produced by different types of immune cells, mainly T-lymphocyte subsets (Fig. 1) and could be classified into three main groups viz Th1 cytokines (TNF, INF-γ, IL12, IL2, etc.), Th2 cytokines (IL4, IL5, IL6, IL13, etc.) and regulatory cytokines (IL10, TGF-β) (Delcenserie et al. 2008). Accumulating evidence has indicated that probiotic strains have the ability to prevent or attenuate allergic diseases by altering Th1 responses (Chuang et al. 2007). Moreover, certain probiotic strains may promote the production of cytokines, including IFNγ, IL12, IL2, TNFα and IL6 from Th1/Th17 cells (Boirivant and Strober 2007), while some of other probiotic strains may promote the production of regulatory cytokines, including IL10 and TGF-β from Tregs (Belkaid and Tarbell 2009; Zhang et al. 2010). It was recently demonstrated that Lact. plantarum, Lact. salivarius and Lactococcus lactis attenuate Th2 response and increase the production of regulatory cytokines in healthy mice in a strain-dependent manner (Smelt et al. 2012). Live and heat-killed lactobacilli affect cytokines production. In a study conducted by Lee et al. (2011), it was found that the levels of INF-γ were strain-dependent, and Lactobacillus gasseri strains and Lact. plantarum strains induced more than the positive control LGG. The production of Th1 cell-specific cytokines (IFN-γ and IL2) increased, while Th2-specific cytokines (IL4 and IL6) decreased in the supernatant of cultured splenocytes, collected from mice fed with probiotic Dahi compared with the other groups (Shalini et al. 2010). The heat-killed bacteria (containing no bacterial metabolites) induced proliferation as well in a strain-dependent manner, suggesting that the surface properties of some of lactobacilli elicit an immune response, as reflected by the release of INF-γ and lymphocyte proliferation (Lee et al. 2011). In a similar study, live and heat-killed LcZ administrated via oral route to BALB/c mice showed a significant effect on sera level of INF-γ, IL12 compared with control group. However, TNF-α was significantly reduced and no change was observed on IL1α level (Ya et al. 2008). Therefore, the immunoregulatory effects of LcZ was ascribed to a skew of the Th1/Th2 balance in favour of Th1-mediated immunity as it has been shown with Lactobacillus brevis SBC8803 (Segawa et al. 2008). Heat-treated Lact. crispatus K-11 did not influence the Th1/Th2 balance as illustrated by no change in the IFN-γ+CD4+ and IL-4+CD4+ cells in the normal C#H/HeN mice (Tobita et al. 2010). However, it reduced allergic symptoms by shifting the Th1/Th2 balance from Th2 dominant state to a Th1 dominant state in type I allergic model mice (Tobita et al. 2009). The metabolites, especially the bioactive peptides released during fermentation by LAB could contribute to the known immunomodulatory effects of probiotic bacteria. The peptidic fraction released in milk fermented by Lactobacillus helveticus R389 was evaluated for its immunomodulatory properties in a BALB/c murine model after E. coli O157:H7 infection. Cytokine profiling revealed stimulation of a Th2 response in mice fed the peptidic fraction, whereas infected controls demonstrated a proinflammatory Th1 response (LeBlanc et al. 2004). The use of Lact. casei Shirota was also able to enhance NK cell activity and this activity was correlated with an IL12 production, another cytokine implicated in NK cells activity (Takeda et al. 2006). This study suggests that probiotics may play a major role in boosting the immunosurveillance of NK cells, thus helping to prevent the development of malignant tumours.
To decipher the mechanism by which probiotics influence the cytokine production, the WTA and LTA of Lact. casei Shirota and Lact. plantarum ATCC 14917 have been shown in vitro to skew IL10/IL12 ratios in macrophages towards IL10 via a TLR2–extracellular signal-regulated kinase (ERK) signalling axis (Kaji et al. 2010). It has been demonstrated also that the absence of Lact. plantarum TA D-alanylation affects its pro- and anti-inflammatory immunomodulatory capacity in murine host cells in vitro. In murine DCs, the absence of D-alanylation substitution did not affect the Lact. plantarum-induced pro-inflammatory TNFα response, whereas a trend towards an increased IL10 response and IL10/TNFα ratio was observed after co-incubation with D-alanylation negative derivative (dltX-D) as compared with wild type, suggesting that the immunomodulatory capacity is reflected by the composition of TA (Smelt et al. 2013). Moreover, in vivo transcriptome studies investigating adult human duodenal responses have suggested that BCL-9, ERK3, jun proto-oncogene (JUN) and poly(ADP-ribose) polymerase (PARP)14 are involved in signalling events induced by LGG consumption, and that signalling via an IFN–signal transducer and activator of transcription (STAT) 4 axis stimulates the production of mainly Th1-type cytokines (van Baarlen et al. 2011).
Effects on immunoglobulin titres
Immunoglobulin (Ig) production occurs during the adaptive immune response. Several studies confirmed that probiotic induced changes in Ig level in serum. IgG is predominately involved in the secondary antibody response, which occurs later following antigen recognition. IgG level in serum of mice fed with LcZ was significantly increased as compared with control group. The mechanism of action is still unknown but it is conceivable that some biological activate peptides released from LcZ may involve in triggering secondary antibodies responses (Ya et al. 2008). Similarly, oral inoculation of recombinant Lact. casei 525 into specific-pathogen-free BALB/c mice resulted in significantly high levels of IgG in serum with prominent IgG1 titres as well as IgG2a and IgG2b titres indicating the use of recombinant Lact. casei 525 for future vaccine development against E. coli ETEC C83919 (Liu et al. 2009). The continuous administration of multistrain probiotics Lact. rhamnosus NBM-01-07-001, Lact. acidophilus NBM-01-07-002, Lact. plantarum NBM-01-07-003, Bifidobacterium longum NBM-01-07-004 and Enterococcus faecium NBM-01-05-001 for 28 days to BALB/c mice orally, showed the same rise effect on the total serum IgA and IgG antibody levels compared with control (Ho et al. 2011). Many other investigators observed that the total serum IgG level increased after administration of lactobacilli to Balb/c mice (Tsai et al. 2008; Ya et al. 2008). The effect on subtypes was also noticed by few authors. In this regard, after oral administration, Lactobacillus johnsonii NCC 533 skewed systemic IgG isotypes towards a greater proportion of IgG1, an isotype that is associated with IL4 induction of B cells and a Th2 predominant immune response. In contrast, Lact. paracasei NCC 2461 induced a greater proportion of IgG2a, which results from IFN-γ stimulation of B cells, and is associated with a Th1 predominant immune response (Ibnou-Zekri et al. 2003). Treatment with orally administered stationary phase Lact. casei ATCC 393 or Lactobacillus murinus CNRZ increased the IgG1/IgG2a ratio without altering the overall amount of systemic IgG. This increased IgG1 response may reflect a higher CD4+ Th2 cell activity in these mice (Maassen et al. 2003). In allergic condition induced by intraperitoneal injection of ovalbumin, mice fed with Lact. helveticus fermented milk and its cell-free supernatant (CFS) showed a less production of total IgE and ovalbumin-specific IgE in the serum, underscoring that the metabolites produced in CFS could be responsible for the observed effect (Kapila et al. 2008).
Effects on lymphocyte proliferation and Peripheral Blood Mononuclear cell (PBMC)
Lymphocyte proliferation is usually determined by using the 3-(4, 5 dimethylthiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT) cell proliferation assay. It is a colorimetric assay that measures mitochondrial function, which serves as an index of live, metabolically active cells. In a recent study, splenic lymphocytes were treated for 72 h with different lactobacilli strains. Some of them have no effect, while others promoted lymphocyte proliferation with Lactobacillus fermentum LA12, and Lact. plantarum CJMA1, CJLP56, CJLP133, CJLP243, CJNR26 and BJ53 being more effective than the LPS positive control, illustrating their potential as immunomodulators (Lee et al. 2011). Probiotic can also show no stimulation in lymphocyte proliferation, but it remains strain specific. The proliferation of lymphocyte from mice spleen in response to standard mitogens (LPS and ConA) in vivo after 15, 18 and 25 days of daily administration of Lact. plantarum showed no variation in a spleen response to LPS, but slight increase in the proliferative response to ConA only after 18 days, followed by a reduction as in control 25 day of daily feeding (Bujalance et al. 2007). This result showed that the stimulation with Lact. plantarum was not continuous throughout the 25 days of feeding. To understand how lactobacilli could induce lymphocyte proliferation, the effect of the structure of TA was investigated. It was observed that in the absence D-alanylation substitution in TA, the generation of Treg cells decreased significantly suggesting that this compound is also necessary for stimulation and proliferation of T cells subset in healthy mice (Smelt et al. 2013).
The PBMC are generally used in probiotic studies to identify probable immunomodulatory characteristics of lactobacilli to obtain additional information on their potential probiotic effects. Drouault-Holowacz et al. (2006) have tested the capacity of lactibiane tolerance strains (a mixture containing 4 × 109 CFU g−1 of viable lyophilized LAB consisting of 12·5% Bifidobacterium lactis LA 303, 12·5% Lact. acidophilus LA 201, 50% Lact. plantarum LA 301 and 25% of Lactobacillus salivarius LA302) to induce the secretion of IL10 and IL12 after 24 h of culture with human PBMC. Individually, Lact. salivarius LA302 and Bif. lactis LA303 strongly induced IL10 more than a mixture, and they both lowly stimulated the production of pro-inflammatory IL12 (Drouault-Holowacz et al. 2006). The variation in immunomodulatory effects between species is even larger than the variation between the strains of the same species. After 1 and 4 days of co-culture of human PBMC with different species of Lact. acidophilus and Lact. plantarum, all tested strains induced differently the production of IL1β, IL10, IFN-γ and TNF-α with Lact. plantarum exhibiting higher induction capacity of IFN-γ, IL12 and TNF-α compared with Lact. acidophilus strains (Vissers et al. 2010). Similarly, live cells of probiotic lactobacilli (Lact. casei Shirota, LGG, Lact. plantarum NCIMB 8826 and Lactobacillus reuteri NCIMB 11951) after individual incubation with human PBMC for 24 h showed an increased production of IL1β, IL6, IL10, TNF-α, granulocyte-macrophage colony-stimulating factor and macrophage inflammatory protein 1α to different extents, but had no effect on the production of IL2, IL4, IL5 or TNF-β (Dong et al. 2012). The reason behind the strain specificity could lay on soluble factors produced by probiotic that could modulate cytokine production by PBMC. To exemplify that, PBMC treated with Lact. reuteri CRL1098 soluble factors (Lr-S) showed a reduced production of TNF-α (Mechoud et al. 2012). It is important to note that the use of PBMC is an artificial screening tool, as under normal healthy conditions, intestinal bacteria do not enter the circulation, and thus cannot communicate directly with such cells because of the ‘firewall’ effect guaranteed by the MLN (Macpherson and Smith 2006).
Effects on macrophage function and leucocyte count
The activation of peritoneal macrophages is one parameter that indicates the stimulation of the systemic immune system but the mechanism used by probiotic strains is still unclear. Several authors observed the increase in the phagocytic activity of peritoneal macrophage after 2, 5 or 8 days of feeding with probiotics cultures (Kapila et al. 2007, 2012; Rajpal and Kansal 2009), even after 60 days (Kapila et al. 2013). Same effects were perceived with the probiotic Dahi culture namely DI (containing Lact. acidophilus, Lact. casei and Lc. lactis ssp. lactis biovar diacetylactis) and DII (containing Lact. casei mixed Dahi culture BD4) regarding the phagocytic activity as well as the increase in lysosomal enzymes activity of peritoneal macrophages (Shalini et al. 2010). In another study, undernourished mice, sacrificed after feeding with the fermented milk containing yoghurt starter cultures and probiotic Lact. casei DN-114-001 (108 CFU ml−1), presented the increase in phagocytic activity of macrophages isolated from peritoneum and spleen (Galdeano et al. 2011).
Leucocytes in blood primarily consist of neutrophils, lymphocytes and monocytes. Differential white blood cell count is an important criterion in detection of infection with viruses and/or bacteria. Few works are showing the effect of probiotic on leucocyte count either differential or total. As they have the Generally Recognized As Safe (GRAS) status, probiotic strains usually do not induce significant changes in leucocyte count in healthy animals compared with control groups. Bujalance et al. (2007), after feeding mice through intragastric route for 25 days continuously with Lact. plantarum at a dose of about 109 viable bacteria in 100 μl of skimmed milk, did not find a difference in total leucocyte count when comparing with control mice who received skimmed milk only. Similarly, no statistical difference was observed in a count of lymphocytes, neutrophils and monocytes after feeding of Lact. acidophilus 381-IL-28 (Luyai 2004).
Another evidence of the systemic influence of the microbiota on systemic immune responses was recently provided. Despite by being contained by mucosal immunity, gut probiotics can impact responses at distal sites. Interestingly, a recent study demonstrated that PG from radio-labelled E. coli could be found in the serum and improved killing of Streptococcus pneumoniae and Staphylococcus aureus by bone-marrow-derived neutrophils in an NOD1-dependent manner (Clarke et al. 2010). Importantly, detection of MAMPs is not limited to the intestinal mucosa but also has been shown to occur in the bone marrow where signalling can alter hematopoiesis (Molloy et al. 2012).
Effect of probiotic lactobacilli in challenge with pathogens
Challenge studies are useful to assess the effectiveness of probiotic cultures. Scientific evidences are showing the ability of probiotics to prevent or protect and cure several pathological conditions with pathogenic bacteria. Results are quite variable depending upon the specie and the concentration of pathogenic/probiotic strain, the pre- and postinfection time period, the age and health status of animals used and many more factors. Different pathogens have been used in animal models for challenge studies with probiotics. The protective effect of Lact. casei DN-114-001 on translocation of Salmonella typhimurium to spleen, liver and large intestine was prominent in mice and illustrated by a significant decrease in Salmonella counts (Galdeano et al. 2011). In a similar study, the level of IgA and IgG in serum and S-IgA was increased only in co-infection group of the probiotic Lactobacillus sp. Dad13 and Salm. typhimurium (Kusumawati et al. 2006). The administration of Lact. casei in the form of fermented milk and nonfermented milk to mice infected with Shigella dysenteria reduced the colonization in the intestine, liver and spleen and increased the concentration of S-IgA in the intestinal fluid of probiotic groups (Kapila et al. 2007). In the work conducted by De Moreno de LeBlanc et al. (2010) investigating the preventive and therapeutic effect of the probiotic Lact. casei CRL 431, results obtained show that 7 days prefeeding Lact. casei CRL 431 decreased the severity of the infection with Salmonella enteritidis serovar Typhimurium, demonstrating that the continuous administration (even after infection) had the best effect. This continuous administration diminished the counts of the pathogen in the intestine as well as its spread outside this organ (De Moreno de LeBlanc et al. 2010). The oral administration of Lact. casei CRL 431 induced variations in the cytokine profile and in the TLRs expression, before and also after the challenge with Salm. typhimurium in mice. The probiotic administration to healthy mice increased the expression of TLR2, TLR4 and TLR9 and improved the production and secretion of TNF-α, IFN-γ and IL10 in the inductor sites of the gut immune response (Peyer's patches) (Castillo et al. 2011). Same authors mentioned that Lact. casei CRL 431 administration increased TLR5 expression after Salmonella infection in groups that received the probiotic strain for 7 days postchallenge (Castillo et al. 2011). The mechanism of protection by probiotic against pathogenic bacteria is quite diverse. Certain probiotic strains (Lact. plantarum, Bifidobacterium infantis) have the capacity to enhance gut barrier function through altering gene expression and distribution of occludin, ZO-1, ZO-2 and different claudin isoforms, therefore strengthening the junction between epithelial cells in the interface between the host and the environment (Ewaschuk et al. 2008; Miyauchi et al. 2012; Mathias et al. 2013; Pruteanu and Shanahan 2013). Probiotics are also known to reduce antibiotic-associated diarrhoea (AAD) and Clostridium difficile-associated diarrhoea (CDAD) risk in a strain-specific manner (Allen et al. 2013; Ouwehand et al. 2014). The Gram-positive (VSL#3) and Gram-negative EcN (E. coli Nissle) probiotic culture could differentially protect against Salmonella dublin infection in vitro using HT29 IEC, illustrating the profound differences existing in the composition of the bacterial cell wall (Mathias and Corthésy 2011). EcN as well as debris and cell extracts induced IL8 secretion from IEC, whereas no such effect was observed following incubation with the VSL#3 (Otte and Podolsky 2004). The latter and its soluble protein(s) increased transepithelial resistance (TER), prevented pathogen-induced decrease in TER, and were shown to stabilize tight junctions and to induce expression of mucins in IEC. By different mechanisms, they both diminished S. dublin-induced cell death (Otte and Podolsky 2004). Others probiotics have been reported to induce secretion of S-IgA (Kapila et al. 2007), cytokines and Ig in serum (Kusumawati et al. 2006), secretion of defensins, mucins, heat shock proteins and antimicrobial peptides from epithelial cells (Schlee et al. 2008; Salzman et al. 2010; Gourbeyre et al. 2011), secretion of bacteriocins (Servin 2004), decreased adhesion of pathogens on epithelial cells (Sun et al. 2007) and stimulation of cytokine-producing cells at Peyer's patches (Castillo et al. 2011).
Use of lactobacilli as a delivery vehicle for antigens in immunization protocols
Food-grade lactobacilli have been safely consumed for centuries by humans in fermented foods. Thus, they are good candidates to develop oral vectors, constituting attractive alternatives to attenuated pathogens, for antigen delivery strategies in immunization. Mucosal surfaces are the primary interaction sites between an organism and its environment and they thus represent the major portal of entry for pathogens. Mucosal immunization is advantageous over other routes of antigen delivery because it can induce both mucosal and systemic immune responses (del Rio et al. 2008). There have been numerous reports of successful immunization with a variety of lactobacilli strains, showing efficient systemic immune response with less collateral side effects than systemic vaccines. The immunopotential of the recombinant strain of Lact. casei capable of producing Beta-lactoglobulin (BLG), one of the major allergens found in cow's milk, was investigated in cow's milk allergy model (Hazebrouck et al. 2009). Intranasal preadministration of the BLG-producing Lact. casei enhanced BLG-specific IgG2a and IgG1 responses in serum, but did not influence BLG-specific IgE production in sensitized mice (Hazebrouck et al. 2009). Unexpectedly, oral preadministration led to a significant inhibition of BLG-specific IgE production, whereas IgG1 and IgG2a responses were not stimulated in sensitized mice (Hazebrouck et al. 2009). These results indicated that the mode of administration of recombinant LAB might be critical for their immunomodulatory properties. Similarly, oral immunization with Lact. plantarum expressing wild-type OspA (the outer surface protein A of Borrelia burgdorferi involve in Lyme disease) induced primarily IgG2a, a Th1-driven immune response, that could be advantageous for vaccination strategies (del Rio et al. 2008). The recombinant lactobacilli can induce specific humoral (protective) and mucosal antibodies and cellular immune response against protective antigens upon nasal administration. Lactobacillus plantarum strain of human origin (NCIMB8826) expressing the 47-kDa fragment C of tetanus toxin (TTFC) was used to elicit immune response by intranasal delivery using mice model (Grangette et al. 2001). The recombinant strain elicited very high systemic immunoglobulin G (IgG1, IgG2b, and IgG2a) responses in serum which correlated to the antigen dose (Grangette et al. 2001). Moreover, significant TTFC-specific mucosal IgA responses were measured in bronchoalveolar lavage fluids, and antigen-specific T cell responses were detected in cervical lymph nodes, both responses being higher in mice receiving a double dose of bacteria (at a 24-h interval) at each administration (Grangette et al. 2001). Few years later, the same Lact. plantarum NCIMB8826 was engineered to express UreB, a Helicobacter pylori urease subunit and deliver it to the gastrointestinal tract. Intragastric immunization elicited UreB-specific antibodies (Corthesy et al. 2005). Moreover, stomach load of Helicobacter felis reduced significantly after challenge in mice fed with the recombinant Lact. plantarum (Corthesy et al. 2005). This showed a successful induction of partial protection against H. felis. Many other recent studies performed with strains of Lact. casei (Poo et al. 2006; Kajikawa et al. 2007; Lee et al. 2010; Kajikawa 2012; Wen et al. 2012b), Lact. acidophilus (Moeini et al. 2010) and Lact. johnsonii (Scheppler et al. 2002) showed the efficiency of lactobacilli as antigen delivery vehicle. However, comparison of the different vaccine delivery and adjuvant systems is difficult for several reasons that include antigen nature, dose and cellular location of the recombinant (cytoplasmic vs bound to the cell membrane vs bound to the cell wall), intrinsic antigenicity and immunomodulatory properties of the bacterial carrier, mouse strain and immunization protocol.
In summary, although specific in vivo or in vitro immunomodulatory properties can be attributed to specific bacteria, probiotic-induced immunomodulation is a complex interplay of different host-microbe interactions. The task remains difficult due to the specificity of their stimulation that could be attributed to the varying composition of MAMPs involved in cross-talk with host immune system, and the plethora of active metabolites produced by probiotic lactobacilli. The puzzle of the precise mechanism is not yet completed but it seems likely that the intestinal epithelial cells, macrophages, dendritic cells (from lamina propria and circulation) and lymphocytes are all differently affected in some way by probiotic strains or metabolites, making the statement of a general mechanism arduous. On the other hand, the use of probiotic lactobacilli strains as delivery vehicle for antigen is an alternative strategy to successfully overcome the disadvantage of inactivated vaccines. The variety of antigens evaluated, the progress made so far in the immune parameters involved in the protective effect of recombinant vaccines and the safety considerations of their application in humans suggest that the use of recombinant lactobacilli as an antigen vehicle represents a tangible option to control infections. Future directions in immunomodulatory studies of lactobacilli should continue to elucidate the effects of single probiotic strain to explore all currently unknown and unappreciated pathways in the host and to complete what is known. The challenge over the next few years will be to develop novel experimental approaches allowing the exploration of the complexes and constantly evolving interactions between probiotic lactobacilli and their hosts.
The authors are thankful to the Director of National Dairy Research Institute (NDRI, Karnal) for providing facilities. T.S. Kemgang thanks the Ministry of External Affairs (Government of India) and the African Union for the award of Doctoral Fellowship.
Conflict of Interest
The authors declare no conflict of financial interest.